In the first funding phase, the Modeling Initiative constructed a number of relational and biophysical models of mechanics and molecular phenomena related to cell migration and started to develop migration related capabilities within the Virtual Cell (VC) software. This activity can be considered as the last step in the reductionism agenda - in silico reconstitution of a simplified motile system using mathematical representation combining biological knowledge and hypotheses, with determination of the consequences of these hypotheses facilitated beyond human reasoning by means of computer-generated numerical calculations. These models and software development enabled the exciting possibility to make a large, critical step in our quantitative understanding of cell migration from the point of view of systems biology. The models will be standardized from the technical point of view, integrated, comprehensive and predictive. A crucial feature of our endeavor, absolutely required for validating such models and using them for hypothesis prediction-test efforts, is that no modeling is undertaken absent direct input from experimental collaborators. We will describe below the mechanism by which this requirement will be consistently met. The Modeling Initiative will investigate migration mechanisms at the systems-level with a long term goal of developing a comprehensive model of cell migration. This model will have a modular character combining deterministic and stochastic components. Our approach is to develop models for each of the component processes that drive cell migration, e.g., development of polarity, protrusion, adhesion, and contraction and rear release, and then integrate them into a comprehensive model. For each of these processes, a 'Process Team'that includes both computational biologists and experimental biologists (in a few cases, these capabilities reside within the same laboratory ), will work together to develop a 'Process Model'capable of capturing dynamic behavior in terms of molecular properties (protein levels, states, locations, and activities). It is through this collaborative team that data will be produced, analyzed, and modeled iteratively with a goal of developing additional data from model predictions and using these data to refine the models. The models for the component processes, e.g., "Process Models", will be implemented in the framework of the Virtual Cell software, to facilitate portability among Teams and accessibility to external investigators in the community. The data will also be made available to the community for their modeling efforts. An ultimate goal is to "stitch" all the Models together to generate a single, unified Cell Migration Model integrating all of the component processes and their regulation. The unified model will have a number of "mechanical" modules, such as protrusion-boundary movement, adhesion-contraction, a number of "biochemical regulation" modules, each representing an individual regulation pathway, and a number of "molecular" modules, for example, an adhesion assembly and disassembly module, an actin assembly and disassembly module, etc (See Figure above). All these modules will be implemented as interacting "boxes" (the best studied ones in the language of differential equations, less studied ones in the language of Boolean or relation networks). The comprehensive model will be "openended". That is, it will be implemented as a distributed application freely accessible over the internet to all members of the scientific community, with the option to change/add models to test biological hypotheses. We will also devise VC tools to deposit, systematize and maintain experimental images on the web, and to perform semi-automatic image and bioinformatics analysis, and a system to compare simulations with data and to generate hypotheses for experimentalists. The comprehensive model will be considered successful if it reproduces observed migration phenotypes and if it is useful as a part of the biological discovery process. The comprehensive model of migration will be data-driven. The crucial quantitative data are the spatiotemporal distributions of molecules essential for cell migration and rates for motility events and their relative strengths. We will collaborate with Signaling and Imaging Initiatives to obtain these data using correlative methods and acute perturbations to visualize actin and adhesion dynamics in moving cells. Specifically, use of CALI, biosensors and caged signaling compounds, fluorescence assays combined with quantitative fluorescent speckle and correlation microscopy will let us probe and determine the spatial and temporal correlations between signaling, mechanical and transport motility events. Our Initiative will assist both by providing automated data acquisition and analysis, as well as utilizing the results in computational models. Imaging and quantitative image analysis will be an integrated part of the modeling. We will have a centralized database where raw image data will be stored, and will utilize image analysis algorithms to extract information that will be used by one of the models. These data bases and tools will be available to the community for their modeling efforts. The models will have parameter fitting front end modules which will interact with the database. These modules will retrieve the data necessary for testing a specific aspect of a model.